Remove stray code alignment (.align 16 and .align 4 directives)

from i386 and x86_64.

ldso/ldso/i386/dl-startup.h

@@ -5,7 +5,6 @@
  */
 __asm__ (
     "	.text\n"
-    "	.align 16\n"
     "	.globl	_start\n"
     "	.type	_start,@function\n"
     "_start:\n"

ldso/ldso/i386/resolve.S

@@ -21,7 +21,6 @@
  */
 
 .text
-.align 4
 
 .globl _dl_linux_resolve
 .type	_dl_linux_resolve,@function

ldso/ldso/x86_64/dl-startup.h

@@ -8,7 +8,6 @@
  */
 __asm__ (
 	"	.text\n"
-	"	.align 16\n"
 	"	.global _start\n"
 	"	.type   _start,%function\n"
 	"_start:\n"

ldso/ldso/x86_64/resolve.S

@@ -26,7 +26,6 @@
 
 .global _dl_linux_resolve
 .type   _dl_linux_resolve,%function
-.align 16
 
 _dl_linux_resolve:
 	subq $56,%rsp

libc/string/sparc/_glibc_inc.h

@@ -6,6 +6,8 @@
 #include <features.h>
 #include <bits/wordsize.h>
 
+/* Is alignment really needed? */
+
 #if __WORDSIZE == 32
 # define ENTRY_ALIGN 4
 #else

libc/string/x86_64/_glibc_inc.h

@@ -6,15 +6,8 @@
 #include <features.h>
 #include <bits/wordsize.h>
 
-#if __WORDSIZE == 32
-# define ENTRY_ALIGN 4
-#else
-# define ENTRY_ALIGN 2
-#endif
-
 #define ENTRY(sym) \
 	.global sym; \
-	.align  ENTRY_ALIGN; \
 	.type   sym,%function; \
 	sym:
 

libc/string/x86_64/memset.S

@@ -71,7 +71,7 @@ ENTRY (memset)
 	jae	11f
 
 	.p2align 4
-3:	/* Copy 64 bytes.  */
+3:	/* Fill 64 bytes.  */
 	mov	%r8,(%rcx)
 	mov	%r8,0x8(%rcx)
 	mov	%r8,0x10(%rcx)
@@ -84,7 +84,7 @@ ENTRY (memset)
 	dec	%rax
 	jne	3b
 
-4:	/* Copy final bytes.  */
+4:	/* Fill final bytes.  */
 	and	$0x3f,%edx
 	mov	%rdx,%rax
 	shr	$0x3,%rax
@@ -107,7 +107,7 @@ ENTRY (memset)
 	jne	8b
 9:
 #if BZERO_P
-	nop
+	nop	/* huh?? */
 #else
 	/* Load result (only if used as memset).  */
 	mov	%rdi,%rax	/* start address of destination is result */
@@ -115,7 +115,7 @@ ENTRY (memset)
 	retq
 
 	.p2align 4
-11:	/* Copy 64 bytes without polluting the cache.  */
+11:	/* Fill 64 bytes without polluting the cache.  */
 	/* We could use	movntdq    %xmm0,(%rcx) here to further
 	   speed up for large cases but let's not use XMM registers.  */
 	movnti	%r8,(%rcx)

libc/string/x86_64/strcat.S

@@ -21,6 +21,7 @@
 
 #include "_glibc_inc.h"
 
+/* Seems to be unrolled too much */
 
 	.text
 ENTRY (BP_SYM (strcat))

libc/string/x86_64/strchr.S

@@ -20,6 +20,7 @@
 
 #include "_glibc_inc.h"
 
+/* Seems to be unrolled too much */
 
 	.text
 ENTRY (BP_SYM (strchr))

libc/string/x86_64/strcpy.S

@@ -20,6 +20,8 @@
 
 #include "_glibc_inc.h"
 
+/* Seems to be unrolled too much */
+
 #ifndef USE_AS_STPCPY
 # define STRCPY strcpy
 #endif

libc/string/x86_64/strcspn.S

@@ -25,6 +25,9 @@
 
 #include "_glibc_inc.h"
 
+/* Seems to be unrolled too much */
+/* (and why testb $0xff, %cl, not testb %cl, %cl?) */
+
 /* BEWARE: `#ifdef strcspn' means that strcspn is redefined as `strpbrk' */
 #define STRPBRK_P (defined strcspn)
 

libc/string/x86_64/strlen.S

@@ -20,6 +20,7 @@
 
 #include "_glibc_inc.h"
 
+/* Seems to be unrolled too much */
 
 	.text
 ENTRY (strlen)

libc/sysdeps/linux/i386/__longjmp.S

@@ -24,7 +24,6 @@
 
 .global __longjmp
 .type   __longjmp,%function
-.align 4
 __longjmp:
 	movl 4(%esp), %ecx	/* User's jmp_buf in %ecx.  */
 	movl 8(%esp), %eax	/* Second argument is return value.  */

libc/sysdeps/linux/i386/bsd-_setjmp.S

@@ -27,7 +27,6 @@
 
 .global _setjmp
 .type   _setjmp,%function
-.align 4
 _setjmp:
 	xorl %eax, %eax
 	movl 4   (%esp), %edx

libc/sysdeps/linux/i386/bsd-setjmp.S

@@ -23,7 +23,6 @@
 
 .global setjmp
 .type   setjmp,%function
-.align 4
 setjmp:
 	movl 4   (%esp), %eax
 	/* Save registers.  */

libc/sysdeps/linux/i386/mmap64.S

@@ -40,7 +40,6 @@
 .text
 .global mmap64
 .type   mmap64,%function
-.align 1<<4
 
 mmap64:
 	/* Save registers.  */

libc/sysdeps/linux/i386/setjmp.S

@@ -24,7 +24,6 @@
 
 .global __sigsetjmp
 .type   __sigsetjmp,%function
-.align 4
 
 __sigsetjmp:
         movl 4   (%esp), %eax

libc/sysdeps/linux/i386/sigaction.c

@@ -145,7 +145,6 @@ libc_hidden_weak(sigaction)
 __asm__						\
   (						\
    ".text\n"					\
-   "	.align 16\n"				\
    "__" #name ":\n"				\
    "	movl $" #syscall ", %eax\n"		\
    "	int  $0x80"				\
@@ -163,7 +162,6 @@ RESTORE (restore_rt, __NR_rt_sigreturn)
 __asm__						\
   (						\
    ".text\n"					\
-   "	.align 8\n"				\
    "__" #name ":\n"				\
    "	popl %eax\n"				\
    "	movl $" #syscall ", %eax\n"		\

libc/sysdeps/linux/i386/syscall.S

@@ -23,7 +23,6 @@
 .text
 .global syscall
 .type   syscall,%function
-.align 4
 syscall:
 	pushl %ebp
 	pushl %edi

libc/sysdeps/linux/i386/vfork.S

@@ -15,7 +15,6 @@
 .global __vfork
 .hidden __vfork
 .type   __vfork,%function
-.align 1<<4
 
 __vfork:
 	popl %ecx

libc/sysdeps/linux/sparc/_math_inc.h

@@ -9,6 +9,8 @@
 #define __ASSEMBLY__
 #include <asm/traps.h>
 
+/* Is alignment really needed? */
+
 #if __WORDSIZE == 32
 # define ENTRY_ALIGN 4
 #else

libc/sysdeps/linux/x86_64/__longjmp.S

@@ -26,7 +26,6 @@
    void __longjmp (__jmp_buf env, int val).  */
 .global __longjmp
 .type   __longjmp,%function
-.align 16
 __longjmp:
 	/* Restore registers.  */
 	movq (JB_RBX*8)(%rdi),%rbx

libc/sysdeps/linux/x86_64/bsd-_setjmp.S

@@ -23,7 +23,6 @@
 
 .global _setjmp
 .type   _setjmp,%function
-.align  16
 _setjmp:
 	xorq %rsi, %rsi
 	jmp __sigsetjmp@PLT

libc/sysdeps/linux/x86_64/bsd-setjmp.S

@@ -23,7 +23,6 @@
 
 .global setjmp
 .type   setjmp,%function
-.align  16
 setjmp:
 	movq $1, %rsi
 	jmp __sigsetjmp@PLT

libc/sysdeps/linux/x86_64/clone.S

@@ -53,7 +53,6 @@
 .text
 .global clone
 .type   clone,%function
-.align 4
 clone:
 	/* Sanity check arguments.  */
 	movq	$-EINVAL,%rax

libc/sysdeps/linux/x86_64/setjmp.S

@@ -23,7 +23,6 @@
 
 .global __sigsetjmp
 .type   __sigsetjmp,%function
-.align 4
 __sigsetjmp:
 	/* Save registers.  */
 	movq %rbx, (JB_RBX*8)(%rdi)

libc/sysdeps/linux/x86_64/sigaction.c

@@ -138,7 +138,6 @@ libc_hidden_weak(sigaction)
 asm						\
   (						\
    ".text\n" \
-   ".align 16\n"				\
    "__" #name ":\n"				\
    "	movq $" #syscall ", %rax\n"		\
    "	syscall\n"				\

libc/sysdeps/linux/x86_64/syscall.S

@@ -26,7 +26,6 @@
 .text
 .globl syscall
 .type	syscall,%function
-.align 16
 syscall:
 	movq %rdi, %rax		/* Syscall number -> rax.  */
 	movq %rsi, %rdi		/* shift arg1 - arg5.  */

libc/sysdeps/linux/x86_64/vfork.S

@@ -32,7 +32,6 @@
 .global __vfork
 .hidden __vfork
 .type	__vfork,%function
-.align 16
 
 __vfork: